A trade-intelligence stack should not be a wall of dashboards. It should be one graph — customs, sanctions, freight, supplier — joined, queryable, defensible. TradeCrest is that graph, and the desk that maintains it.
For more than a decade the cross-border trade intelligence market has been divided between two unsatisfactory species of vendor. On one side, the enterprise compliance platform: slow to integrate, expensive to renew, blind to anything that does not fit the original schema, and built around the assumption that the customer will adapt their workflow to the product rather than the other way about. On the other, the narrow tool — a sanctions screen, a customs lookup, a freight tracker — each useful in its own lane, each adding, in aggregate, to the same problem the customer started with. The data does not join.
The TradeCrest premise is that the value is in the join. Customs data is not interesting on its own. Sanctions data is not interesting on its own. Freight telemetry is not interesting on its own. The supplier that ships through Felixstowe under one company name and through Hamburg under another, financed by a bank whose ultimate beneficial owner is one hop from a designated entity — that supplier is interesting, and the only way you see them is if all four data classes sit on the same graph, indexed by the same entity-resolution model, queried through the same surface.
That graph is the product. The three surfaces — CrestLens for analysts, Signal Routes for lane-risk forecasting, CrestAPI for in-house data teams — are three windows onto the same model, calibrated against the same evidence, versioned to the same release cadence. There is no premium tier. There is no different evidence standard for paying versus non-paying entities. The graph is the graph.
Entity resolution is the difference between a customs database and a customs graph. Without it, you have a wall of strings. With it, you have a map of flows.
— Lane Risk Methodology, v3.2
The trade-intelligence workspace for compliance officers, trade-finance underwriters and global sourcing leads. Search by entity, by HS code, by lane, by counterparty bank. One queryable analyst's desk, not five disconnected tools.
Read on CrestLensLive lane-level disruption forecasts. Eight inputs, one decision-ready number per lane per week, with explainable contributing factors. Calibration is published; the methodology is versioned and signed.
Read on Signal RoutesDirect access to the trade graph for in-house data teams. Four endpoints, stable contract, predictable schema evolution, OAuth 2.0 with hardware tokens for any write path. Versioned alongside the graph, not the marketing roadmap.
Read on CrestAPIAll figures are calibrated production numbers from the current release. Where a figure is a target rather than a measurement we say so explicitly; where it is a measurement we publish the test set and the calibration window in the Library. We do not report curated benchmark numbers.
Customs declarations are the most underused goldmine in global trade. We explain how modern entity resolution, schema harmonisation and graph methods are turning a wall of strings into a map of flows — and what compliance teams can actually do with that map.
Apr MMXXVI · 2,900 words · Customs & AIThe UK OFSI Q1 update tightened beneficial-ownership thresholds and added 412 entities. We map the practical impact on commodity traders, shipping lines and trade-finance underwriters — and explain why screening accuracy collapsed for many firms overnight.
Mar MMXXVI · 2,600 words · SanctionsDisruption forecasting is hard not because we lack data but because we have too much. We walk through the eight raw inputs behind every Signal Routes score, the ensemble that fuses them, and the calibration discipline that makes the number actionable rather than decorative.
Feb MMXXVI · 3,100 words · Lane RiskA TradeCrest briefing is a 45-minute working session with the analyst desk. We come prepared with your portfolio against the graph: the entities your existing screening misses, the lanes your insurer has not yet priced, the supplier chains that are one hop closer to a designated party than your last quarterly review suggested. There is no slideware. There is no qualification call. The address below is real, the inbox is monitored, and the reply rarely takes more than a working day.
We are deliberately small. Demand for briefings is, at the time of writing, larger than the desk; we operate a brief queue rather than a sales pipeline, and we will tell you on the first reply where you sit in it. If we cannot help, we will say so on that first reply and refer you to someone who can. We have done both.